• Today in AI
  • Posts
  • Manus AI: The $500m Chinese AI startup that’s winning over the U.S.

Manus AI: The $500m Chinese AI startup that’s winning over the U.S.

🤖 Discover the chinese AI startup that just got funded by American investors at a $500 million valuation

In partnership with

(7 minutes)

Hi 👋, this is the Today in AI Newsletter: The weekly newsletter bringing you one step closer to building your own startup.

We analyze a cool, industry-shaping AI startup every week, with a full breakdown of what they do, how they make money, how much they’ve raised, and the opportunity ahead. 

Let’s get to the good stuff in this email: 

  • 🧱 This chinese startup’s AI agent can help you build anything in minutes

  • 🚀 They hit 2 million people on their waitlist 7 days from launch

  • 💰 They’ve just raised a $75M Series B at a $500M valuation led by Benchmark

p.s. stay till the end to hear the founders philosophy on what it takes to build the most powerful AI

So what’s the startup and who are the founders behind it? Here’s the story of Manus AI 📈

Manus AI was founded by Xiao Hong and Yichao Ji in 2022 with the vision to build a thinking partner that helps people create, explore, and express ideas at the speed of thought itself. ⚡

At its core, Manus AI is a universal task execution agent. Not just a chatbot, not just an LLM wrapper, but a system that can plan, reason, and act across the web to complete multi-step objectives autonomously.

Let’s break that down.

Where most AI tools require prompt-by-prompt interaction (like ChatGPT or Claude), Manus flips the model. ⚙️ Instead of chatting, you give it a goal, something like:

  • 🧾 “Find and compare three AI legal tools for startups.”

  • 🗺️ “Create a 3-day itinerary in Kyoto under $600.”

  • 📊 “Draft a pitch deck for a healthtech company targeting the NHS.”

Then Manus goes to work. Behind the scenes, it does something powerful:

  1.  Manus breaks your request into smaller steps, like researching, comparing, and writing, instead of doing everything at once. ✍️

  2. It picks the best AI model for each step - sometimes it uses a big model like GPT-4, and sometimes a faster one for simple tasks. 🤖

  3. Manus can browse the internet, click on websites, gather information, and even fill out forms - like a smart assistant that knows how to use Google. 🔍

  4. It remembers what you’ve asked before, so you can make changes or keep working without starting over. 🧠

  5. You just type what you need in plain language, no coding, no special setup. Manus figures out how to do it for you. 💬

It’s not just about replacing humans. It’s about amplifying them - letting anyone access leverage normally reserved for teams, VAs, or expensive SaaS stacks. ⚡

That’s the core bet Xiao is making: not that AI will replace people, but that it will make anyone as capable as a team of five.

Manus AI currently operates as an invite-only tool, with a limited number of users at a time. 😎

They’ve had millions of people join the waitlist, which has prompted a black market for Manus AI invite codes selling for as high as $7,000… 💸

Backstory 👀

Before Manus AI, before agents were even a thing, Xiao Hong was building mini productivity apps for Chinese college students.

He was born and raised in Hubei Province, and by the time he enrolled at Huazhong University of Science and Technology (HUST), one of China's top engineering schools, he was already writing code and experimenting with automation tools. 🧑‍💻

While still a student, Xiao noticed a recurring problem: students at Chinese universities were juggling endless administrative tasks - checking schedules, submitting forms, navigating clunky government portals. It was tedious, repetitive, and boring work. 🥱

Instead of complaining, he started building. 🏛️

His first breakout product was Yiban Assistant, a WeChat-based tool that automatically synced class timetables and reminders for university students. 🧑‍🎓

 It was followed by Weiban Assistant, which helped automate daily health check-ins required by schools during COVID. 📲These weren't flashy AI products. They were tools that just worked - simple, efficient, and incredibly sticky.

Word spread quickly. By 2020, over 2 million students across China were using Xiao’s tools and investors took notice. 📈

He raised early funding from Tencent and ZhenFund, two of the most powerful backers in the Chinese tech ecosystem. 💼 

Soon after, his startup was acquired by a unicorn education company. 💰

This gave Xiao his first major exit 💎- and more importantly, the confidence and capital to build globally.

That’s when things got interesting…

Want to make some extra $$?

100 Genius Side Hustle Ideas

Don't wait. Sign up for The Hustle to unlock our side hustle database. Unlike generic "start a blog" advice, we've curated 100 actual business ideas with real earning potential, startup costs, and time requirements. Join 1.5M professionals getting smarter about business daily and launch your next money-making venture.

From China to going Global 🌍

In 2022, Xiao came back with a new company: Butterfly Effect Tech, this time with a mission to build AI tools for a global audience.

China’s domestic market was crowded, competitive, and heavily regulated. 📄

Xiao saw an opportunity 💡: most of the world had access to ChatGPT, but very few people were building practical, consumer-grade tools around these models.

That led to Monica, a Chrome extension designed to be your everyday AI assistant. 

It could summarize web pages, translate articles, rewrite emails, answer questions, and more - all in one clean, lightweight interface. 📚

Monica didn’t try to be everything. It didn’t brag about its model size or token context windows. It simply did a few tasks really well. People loved it. 😆

They went on to raise a $10 million Series A from Tencent and Sequoia Capital China (now HSG) in January 2023. 🤩

Without much marketing, Monica quietly exploded in popularity, especially among non-English speaking users in Latin America, Eastern Europe, and Southeast Asia

At one point, it ranked among the top productivity extensions in the Chrome Web Store, reaching over 10 million users and even turning a profit through premium upgrades. 💸

ByteDance (Owner of TikTok) even attempted to buy the startup for $30 million, but Xiao refused the offer. 💃

Monica was Xiao’s proof of concept: a consumer AI layer that could sit on top of LLMs and actually solve real user problems.

But he wasn’t done yet.

The Next Evolution: Manus AI 🤖

In March 2025, Xiao and his team launched Manus AI, a more ambitious product that builds on everything they learned from Monica.

If Monica was your sidekick, Manus is your full-time operator. 🧑‍💼

It’s an autonomous agent that can take a multi-step task - say, researching vendors, planning a trip, or drafting legal documents - and execute it without constant input. You give it the goal, and it figures out how to achieve it. 💫

The tech isn’t new. AutoGPT and similar agents have existed since 2023. But they’ve been slow, buggy, or confined to the terminal. 

Manus is model-agnostic, it works with GPT-4, Claude, and Qwen (Alibaba’s model) depending on the task - but it hides all that complexity behind a single click. ⚙️

 No setup, no configurations, no headaches. Just results.

Within 7 days of launch, Manus AI had 2 million people on their waitlist… 📈

🔥Within weeks, the product caught fire, especially in China’s tech circles. Manus outscored OpenAI’s Deep Research on the GAIA benchmark (86.5% vs. 80%)...

📸TikTok creators began posting “I let Manus run my business for a day” videos. 

ProductHunt upvotes SURGED. 🔝On Chinese platforms like Zhihu and Xiaohongshu, Manus became a top-trending topic, often compared to OpenAI’s “yet-to-be-released” super agents.

Unlike most AI startups, Manus wasn’t pitching to enterprises or demoing research papers. It was shipping features, fast - and most importantly, gaining real users. 👱

The Benchmark x Manus Controversy 📉

💰On April 25th 2025, Manus AI raised a $75M Series B at a $500 million valuation led by Benchmark, but the celebration was short-lived.

🕵️Weeks later, the U.S. Treasury began reviewing the deal. The reason? Manus is a Chinese-founded AI startup, and the investment may fall under new 2023 U.S. restrictions on funding Chinese tech companies.

Benchmark insists the deal is fully regulation compliant. Their lawyers argue Manus doesn’t build its own AI models, it builds apps on top of open-source ones. 🧑‍⚖️

The company is also legally incorporated in the Cayman Islands, a common workaround used by Chinese firms like Alibaba to raise global capital. 🏖️

Still, the investment drew fire online. Delian Asparouhov from Founders Fund mocked Benchmark’s move on X, posting: “wow, actions have consequences?” - a shot at what some see as VC firms ignoring national security concerns for the sake of returns.

So far, Benchmark, Manus, and the Treasury have all stayed quiet. But, this case could set the standard for how U.S. regulators treat Chinese AI startups operating in a grey area.

Xiao’s Philosophy on AI 🧠

If you ask Xiao what makes a great AI product, he won’t talk about models, inference speed, or vector databases. He’ll tell you this:

“The most powerful AI is the one that actually gets used. Not the one with the best model, but the one that fits into your life.”

His approach is anti-hype. He believes the best ideas don’t need explanation - they just need a good interface.

Where others are obsessed with building the next foundational model, Xiao is more interested in AI packaging, designing wrappers that deliver value fast. 💨

It’s the same thinking that made TikTok beat YouTube Shorts. Or why the iPhone beat Nokia. It wasn’t the hardware. It was the experience.

📲Manus is built in that image: not to replace OpenAI or Claude, but to ride their wave and be the fastest layer of value creation on top.

Before you go 👋

Thank you for reading.

Know someone who'd love this newsletter? Share the love.🌟